No9_NatalieDoerr.qxd 8/3/10 1:54 PM Page 1

Corella, 2010, 34(1): 14-16

ANNUAL SURVIVAL RATES OF BOWER-OWNING MALE GREAT Ptilonorhynchus nuchalis IN TOWNSVILLE, QUEENSLAND

NATALIE R. DOERR

Department of Ecology, Evolution and Marine Biology University of California, Santa Barbara, Santa Barbara, CA 93106. E-mail: [email protected]

Received: 23 February 2009

This study documents the annual survival rate of bower-owning male Great Bowerbirds Ptilonorhynchus nuchalis in Townsville, Queensland. The mean survival rate of bower-owning males was 84 per cent from 2003 to 2007. The estimate of survival is slightly lower than that recorded for bower-owning males of three other species (range: 90 – 93%). Lower estimates could be a consequence of dispersion from the study area, particularly of younger . Lower estimates could also have occurred because Townsville is a suburban environment containing hazards that bowerbirds do not experience under natural conditions. Over the course of the study, one male drowned in a swimming pool and one appeared to have died due to predation. At least eight unbanded individuals (sex and status unknown) were struck by vehicles on the study site.

INTRODUCTION areas used by pedestrians, such as footpaths and car parks. Males were captured at their bowers using mistnets or spring- Males of most species (Ptilonorhynchidae) are loaded mammal traps baited with fresh fruit and bower promiscuous and build bowers to attract females with whom decorations. A metal band issued from the Australian and they mate. Several studies have shown that bowerbirds exhibit Bat Banding Scheme, as well as a unique combination of three high annual survival rates. Though data for females are limited, colour bands were placed on each male’s leg. One colour band the survival rates for bower-owning males of three species was placed above the metal band, and the other two were placed ranged from 90 to 93 per cent (reviewed in Frith and Frith on the opposite leg. 2004). However, additional studies are needed to determine whether high survival rates are a family-wide trend in the Once a male was banded, I returned to the bower on a Ptilonorhynchidae. regular basis to determine whether the banded individual was the owner. At some bowers, I monitored males using a motion- Great Bowerbirds Ptilonorhynchus nuchalis live throughout activated camcorder system, which recorded all activity at northern Australia and build one of the largest and most well- bowers for the majority of the breeding season (September decorated bowers of the Australian bowerbird species. Structures through mid-December). At other bowers, I conducted one- to can exceed one metre in length and are often decorated with more four-hour observation sessions at least twice each breeding than 1,000 objects, including rocks, bones, shells, leaves and fruit season to determine the identity of the bower owner. In (Frith and Frith 2004). Great Bowerbirds are also one of the few addition, I visited most bowers every 10 to 14 days to count bowerbird species that have successfully adapted to human bower decorations, and I noted the identity of individuals development. In Townsville, males frequently build bowers in present during those times. Following Madden et al. (2004), I populated areas, such as parks, school grounds and private defined bower owners as males that were present more than 75 residences (Frith et al. 1996; Doerr 2008); however, this close per cent of the time for which I recorded at least one bird association with humans could affect their survival. present. All presumed bower owners regularly maintained and displayed at their bowers. No studies to date have documented survival rates in any population of this species. This study reports the annual Male bowerbirds exhibit high fidelity to their bower sites survival rates of bower-owning male Great Bowerbirds in (Frith and Frith 2004), and male Great Bowerbirds in Townsville Townsville from 2003 to 2007. typically build new bowers within 33 ± 43 metres (mean ± sd; METHODS range: 1–127 m) of the bowers they built the previous breeding season (Doerr 2008). Therefore, I observed whether male In 2000, researchers at James Cook University began bower-owners returned to their bower sites in successive years, banding Great Bowerbirds in Townsville, Queensland (19°19'S, and I assumed that the bower owner had died if I could not 146°46'E), and I banded and monitored bower-owning males in relocate the male or his bower anywhere on the study site, or on Townsville from 2003 to 2007. Birds were banded at three lands immediately adjacent to the study site, during the adjacent areas encompassing 15 square kilometres: the James following breeding season. Previous studies of survivorship in Cook University campus, the Lavarack Barracks military base, bowerbirds have made the same assumption (Borgia 1993). and the suburb of Annandale. Roads, such as the Bruce Because new bowers were built over the course of the study Highway, intersected all three study areas. Males built their while a few established bowers were abandoned, the number of bowers underneath bushes, and many bowers were adjacent to males that I monitored differs slightly between years.

14 No9_NatalieDoerr.qxd 8/3/10 1:54 PM Page 2

March 2010 N.R. Doerr: Annual survival rates of bower-owning male Great Bowerbirds in Townsville, Queensland 15

To calculate survival rates for each year, I used the equation: was in his first year of bower ownership. A bower-owning male at James Cook University (years of bower ownership unknown) Si = 100 (bi / ai-1) was probably predated: I observed the male at his bower on 14 September 2004; found a mass of bowerbird feathers at his where: S represents survival to year i; a is the number of banded i bower the following morning; and never again saw the owner. birds present in year i-1; and b is the number present the Though I did not know the cause of death for any other bower following breeding season. I performed this calculation owners, I found the bodies of at least eight bowerbirds along the separately for each pair of study years (2003 to 2004, 2004 to side of the Bruce Highway, which intersects the study area. All 2005, etc.). To calculate a single value for the mean annual individuals appeared to have been struck by vehicles. The survival rate over the four years of the study, I used the equation: skeletal remains of a banded, bower-owning male were found in x x a car park at Lavarack Barracks in March 2007 (C. Gould pers. S = 100 (∑ i=1 bi+1 / ∑ i =1 ai) (1) comm.), but the cause of death was unknown. This male had where: a is the number of banded birds present in year i; b is the owned a bower site for at least seven years. number still present the following breeding season; and x is the number of study years (Nicholls and Woinarski 1988). This DISCUSSION equation has been used to estimate survival rates of bower- Like other bowerbird species, bower-owning male Great owning male Tooth-billed Bowerbirds Scenopoeetes dentirostris Bowerbirds in Townsville exhibited high annual survivorship. (Frith and Frith 1995). Because males in their first year of bower The mean survival rate over the four years of the study was 84 per ownership may be more likely to die or disperse from the study cent (sd = 0.034). In previous studies, survival rates of bower- area than more established males (Borgia and Gore 1986; Frith owning males were 90 per cent in Tooth-billed Bowerbirds; 91 and Frith 2004), I also used this equation to create separate per cent in Golden Bowerbirds newtoniana; and 93 survival estimates for males in their first year of bower ownership per cent in Satin Bowerbirds Ptilonorhynchus violaceus (Borgia and males that had been bower owners for two or more years. A 1993, reviewed in Frith and Frith 2004). A study of male Regent value for ‘years of bower ownership’ could not be assigned to Bowerbirds chrysocephalus that included both bower- nine males during the first year in which they were monitored owners and non-bower-owners found annual survival rates of 74 because it was unknown whether these males had been at their per cent (all ages) and 87 per cent (adults only) (Lenz 1999). By bowers for one or more years; these data points were not included the end of my study, seven males had owned bower sites for at in the analysis of survival rates according to years of ownership. least seven years. Longer-term studies in other species have Finally, I performed a 2 x 4 Contingency Table analysis to recorded the same individual at a bower site for up to 20 years determine whether survival rates differed between study years. (Frith and Frith 2004). RESULTS Given the high population density of Great Bowerbirds in The mean annual survival rate over the four years of the Townsville, and the year-round availability of food and water study was 84 per cent (range: 74 – 93%; Table 1). Seven of the (Frith et al. 1996; Doerr unpub. data), it is somewhat surprising 13 males that were bower owners in 2000 were still owners in that survival rates were not higher. The estimate of survival rate 2007 – a minimum of seven years of ownership. Survival rates is similar to that of other species, but it is also the lowest did not differ significantly between study years (2 x 4 recorded for bower-owning males. It is possible that survival Contingency Table, χ2 = 4.35, df = 3, P < 0.23). Males that rates were relatively low because some males left the study site were in their first year of bower ownership had a survival rate to build bowers elsewhere. Although it is uncommon for of 61 per cent, while males that had been bower owners for two established males to abandon their bower sites, even in areas or more years had a survival rate of 89 per cent (Table 2). with considerable disturbance from humans (Frith et al. 1996; Frith and Frith 2004), males in their first year of ownership On 19 September 2003, the staff at Lavarack Barracks found frequently experience bower destruction from their neighbors, one bower-owning male dead in a swimming pool. This male which may cause them to abandon their bowers (Borgia and

TABLE 1 Survival rates of bower-owning male Great Bowerbirds in Townsville from 2003 to 2007. Males that disappeared from their bowers were assumed to have died. For each column, deaths within breeding season = males that disappeared from their bowers between 1 September – 15 December of year 1. Deaths outside of breeding season = males that disappeared between 16 December – 31 August of year 2.

2003-2004 2004-2005 2005-2006 2006-2007

Deaths within breeding season 1 1 1 0 Deaths outside of breeding season 5 4 1 7 Number of bowers monitored 28 34 30 27 % Survival rate (mean ± sd) 79 ± 8 85 ± 6 93 ± 5 74 ± 9 No9_NatalieDoerr.qxd 8/3/10 1:54 PM Page 3

16 N.R. Doerr: Annual survival rates of bower-owning male Great Bowerbirds in Townsville, Queensland Corella 34(1)

TABLE 2

Survival rates of male Great Bowerbirds according to years of bower ownership – see Methods (1). For each column: 'Number present, season 1' = males that were present at bowers during the breeding season of year 1 (1 September to 15 December). Some males were not present for the full breeding season, but they were still included in this tally; 'Number present, season 2' = males that returned to their bower sites the following breeding season. Males that did not return were assumed to have died.

First year Two or more years of ownership of ownership 2003- 2004- 2005- 2006- 2003- 2004- 2005- 2006- 2004 2005 2006 2007 2004 2005 2006 2007

Number present, season 1 771318222923

Number present, season 2 550114222818

Survival rate 61% 89%

Gore 1986). Seven of the 20 males that disappeared over the survivorship found in other bowerbird species. However, future course of this study were in their first year of ownership, and studies documenting survival rates in more natural populations three experienced moderate to severe bower destruction (Doerr of the Great Bowerbird are needed to determine whether unpub. data). If these individuals left the study area in search of Townsville males experience lower survival rates than do other a site where they would experience less interference, then this populations and species. could have biased the estimate of survivorship downwards. Indeed, when survival rates were calculated separately for ACKNOWLEDGEMENTS different age classes, it was found that males in their first year I thank Drs. John Endler and Lainy Day for initiating the project to of bower ownership had a survival rate of 61 per cent, while band Great Bowerbirds; Dr. David Westcott for supervising my males that had been bower owners for two or more years had a banding activities; and Rhona Barr, Rachel Hansen and Itzel Zamora survival rate of 89 per cent. Although the lower apparent Vilchis for assistance in the field. Thanks to Alan McManus for survival rate of first-year males could have been the result of permission to work at Lavarack Barracks. Clifford Frith and Ken dispersion from the study site, younger males may also Rogers provided helpful comments on the manuscript. This study was experience increased mortality (Newton 1989), so additional carried out with permission from the Australian Bird and Bat Banding studies are needed to determine whether first-year males that Scheme, Queensland Environmental Protection Agency, Townsville fail to return the following season have died or dispersed. City Council, and the James Cook University Ethics Committee. Previous studies in Tooth-billed and Golden Bowerbirds have REFERENCES documented survival rates for bower-owning males of all ages (reviewed in Frith and Frith 2004), but Borgia (1993) estimated Borgia, G. (1993). The cost of display in the non-resource-based the survival rate of Satin Bowerbirds based on males that owned mating system of the . The American Naturalist 141: 729–743. ‘permanent’ bower sites. If his estimate excluded first-year Borgia, G. and Gore, M. A. (1986). Feather stealing in the satin males that did not reappear the following year, then this could bowerbird (Ptilonorhynchus violaceus): male competition and the have led to higher survival estimates than this study. quality of display. Behaviour 34: 727–738. Doerr, N. R. (2008). Male-male competition and the cost of display in Alternatively, the results could indicate that male Great the great bowerbird (Chalmydera nuchalis). PhD Thesis, Bowerbirds within Townsville experience a slightly lower University of California, Santa Barbara, CA. survival rate than other species. Though Townsville’s year- Frith, C. B. and Frith, D. W. (1995). Court site constancy, dispersion round availability of food and water may enhance survivorship, and court ownership in the male tooth-billed bowerbird, the suburban environment may contain hazards that bowerbirds Scenopoeetes dentirostris (Ptilonorhynchidae). Emu 95: 84–98. Frith, C. B. and Frith, D. W. (2004). ‘The Bowerbirds: are not exposed to under natural conditions. Townsville males Ptilonorhynchidae’. (Oxford University Press: Oxford.) often build bowers in bushes along the side of roads and swoop Frith, C. B., Frith, D. W. and Wieneke, J. (1996). Dispersion, size and low to the ground when approaching their bowers (Doerr pers. orientation of bowers of the great bowerbird nuchalis obs.). This may increase the likelihood that they will be killed (Ptilonorhynchidae) in Townsville City, Tropical Queensland. by a vehicle. Over the course of the study, bodies of at least Corella 20: 45–55. eight Great Bowerbirds were found that appeared to have been Lenz, N. (1999). Evolutionary ecology of the Regent Bowerbird struck by vehicles on the study site, though none were the Sericulus chrysocephalus. Ökologie der Vögel 22, Supplement, 1–200. bower-owning males included in this study. An earlier study Madden, J. R., Endler, J. A. and Jury, F. (2004). Morphological signals conducted at the same location in Townsville found one banded of sex and status in Spotted Bowerbirds. Emu 104: 1–10. individual (sex unknown) struck by a car (Day pers. comm.). Newton, I. (1989). ‘Lifetime Reproduction in Birds’. (Academic Press: London.) Annual survival rate of bower-owning male Great Nicholls, D. G. and Woinarski, J. C. Z. (1988). Longevity of Pied Bowerbirds is generally consistent with the pattern of high Currawongs at Timbertop, Victoria. Corella 12: 43–47.